6201
Macromolecules 1994,27, 6201-6206
Photoresponsive Behavior of Azobenzene-Based (Meth)acrylic (Co)polymers in Thin Films Henk J. Haitjema, Gijs L. von Morgen, Y. Yong Tan, and Ger Challa’ Laboratory of Polymer Chemistry, University of Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands Received February 8, 1994; Revised Manuscript Received July 15, 1994’ ABSTRACT The reversible photoisomerization and the thermal isomerization of azobenzene-based (A2.b.) groups covalently bound to (meth)acrylic (co)polymerswere investigated in thin films. For the amorphous polymers it was found that a broad range of the thermal cis trans isomerization rates could be obtained by varying the substituent, spacer, and copolymer composition. The liquid crystalline (LC) polymers showed an interesting behavior of the orientation of the stable trans state when exposed to visible light, during which the orientation of the Az.b. groups changed from more or less random to more perpendicular (photoselection process). This change of orientation of the Az.b. groups may be suitable for optical data storage. In principle, the orientation change is permanent, but it can be easily erased by heating the LC polymer above its isotropization temperature, on which the original random orientation is largely restored.
-
1. Introduction Optical data storage materials are mainly based on organic molecules. To be suited for use as an optical data storage material, molecules have to be photochromic; i.e., they must be capable of being reversibly switched between two spectroscopically distinct forms by use of light.’ The photochrome that we have chosen is a substituted azobenzene, which shows a cis-trans isomerization as illustrated in Figure 1. By irradiation with light of wavelength XI or hz the geometric configuration of the azo bond in azobenzenebased compounds can be reversibly switched from trans to cis. The two states have distinct absorption spectra. This feature would allow reversible storage of data on the basis of cis and trans states with the aid of XI and Xz. However, the cis state is thermodynamically unstable with respect to the trans state; therefore a thermal relaxation process occurs in the dark (at room temperature) denoted in Figure 1as A. This thermal back-reaction strongly limits the lifetime of the cis state, and it is influenced by the substituents R1 and Rz, catalysts, temperature, and the environment (solvent polarity, type of polymer matrix) (e.g., refs 2-7). An important factor which strongly affects the rates of both the thermal isomerizationand the photoisomerization of Az.b. compounds is incorporation in a polymer matrix, either as a low molecular dispersion or incorporated in the main chain or the side chain of a polymer. The effect is dependent on the free volume distribution in the polymer matrix.617 If the volume in which the Az.b. group is located is smaller than the volume that is necessary for the trans cis photoisomerization to occur, its rate decreases because of a lower quantum yield and leads to a lower content of cis in the photoequilibrium. On the contrary, the thermal isomerization was accelerated and multiexponential decay kinetics was found. The explanation for the deviating kinetics in the thermal isomerization is that it is probably caused by the individual characteristics of an Az.b. group/free volume combination. If sufficient free volume is available, normal first-order decay is found; if not, a series of first-order decays is found.’ Two methods using the trans cis isomerization have been developed for use in optical data recording. The
-
-
* Author to whom correspondence should be addressed.
Abstract published in Advance ACS Abstracts, September 1, 1994.
*I
trans
cis
Rp
Figure 1. Cis-trans isomerization of Az.b. compounds (for azobenzene R1 = Rz = H).
first system is based on a photoselection process, introduced by Wendorff et ala8 and also used by other investigators (e.g., refs 9 and 10). It involves a change of the orientation of the trans azobenzene side groups with the aid of linearly (for orientation) polarized light. Trans molecules oriented parallel to a substrate but perpendicular to the plane of incoming polarized light are almost inert for photoisomerization and will therefore not be converted to cis. If a trans molecule is converted to cis, its original direction of orientation is lost. The direction of orientation of the trans after thermal back-reaction from cis is random; however, trans molecules perpendicular to the plane of polarization of the incoming light again are not available for photoisomerization. So after some irradiation time mainly perpendicular oriented (with respect to the plane of polarization of the writing beam) trans azo groups remain. Rapid thermal isomerization is essential for this system, since a large number of isomerization steps are necessary for the reorientation. In the second system the trans cis isomerization of the azo compounds is used to influence a LC phase (e.g., refs 11and 12). Now the photoisomerization of trans to cis destroys or changes a surrounding LC phase which is heated close to the isotrspization temperature (Ti), due to the destabilization caused by the large geometrical difference between the bent configuration of the cis state (cis does not “fit”) and the rodlike configuration of the trans state. This change in ordering after irradiation can be retained by cooling below Tr The written data (irradiated areas) can be read with light if the sample is placed between crossed polarizers, and it can be erased by heating the LC polymer above Ti and quenching below Tg (if desired, in the presence of a magnetic or an electric field). In this study the influence of the (co)polymer composition, substituent, spacer, and the addition of complemen-
0024-929119412227-6201$04.50/0 0 1994 American Chemical Society
-
Macromolecules, Vol. 27, No. 21, 1994
6202 Haitjema et al. Table 1. Survey of the Structure and the Results for the Investigated Polymers substituent spacer copolymer polymer h,(t) tl/.2* code I I1 I11 IV V VI VI1
X H H H H H CH3 CH3
Y N(CH3)z N(CH3)z N(CH3)z
n 0 0 6 C02Et 0 N(CH3)z 6 6 COzH COzEt 6
m
p
q
0
0 0.5
1 0.5 0.2 1
0 1 0 1 1 1
0.8 0 0 0 0
1
I
additions (nm) (min) 405 120 -405 110 405 60 325 2700 405 a 360 550
1
360
._
n
a These LC polymers show a peculiar photoresponsive behavior and are discussed in section 4.2. t1p denotes the half-life time of
c1s. X
X 1
-
---CH2-C-dCHp-
c
’p
C - - t c
q N N -
- Y
Figure 2. General structure of the (co)polymers. tary polymers on the thermal isomerization will be discussed, because it has a large influence on the time needed for the writing process. The polymer systems that were studied (see Table 1) possess the general structure presented in Figure 2. An azobenzene group with substituent Y (C02H, N(CH3)2, or C02Et) is bound to a (methlacrylate (co)polymer, with methyl acrylate as comonomer for the acrylates and methyl methacrylate for the methacrylates, through a -(CH2),- spacer with n = 0, 6 , or 11. The amorphous polymers I (Y = N(CH3)z; X = H; n = 0, m = 0; p / q = O / l ) , I1 (Y = N(CH&; X = H; n = 0, m = 0; p / q = 0.5/0.5), I11 (Y = N(CHd2; X = H; n = 6, m = 1; p / q = 0.8/0.2),IV (Y = C02Et; X = H; n = 0, m = 0;p / q = O / l ) , and VI (Y = C02H; X = CH3; n = 6, m = 1;p / q = 0/1) were investigated to establish the dependence of the rate constant of the thermal cis trans back-reaction K ( Tisom) on substituent, spacer length, and copolymer composition. After this the LC polymer V (Y = N(CHd2; X = H; n = 6, m = 1;p / q = 0/1) was investigated in detail.
-
2. Experimental Section
Materials. The synthesis and the characterization of the Az.b. monomers and (co)polymers are described e1~ewhere.l~The solvents used, 1,2-dichloroethane (Merck, Germany),p-dioxane (Janssen Chimica,Belgium),and EtOH (Merck, Germany),were of p.a. quality and used without purification. DMF (Janssen Chimica, Belgium) was distilled before use and stored over 4 A molecular sieves. Polymers I, 11, 111, IV, V, and VI1 were dissolved in 1,2-dichloroethane, and polymer VI was dissolved in DMF. Spin-Coating. Thin polymer films were prepared by spincoating polymer solutionswith aconcentration of 1w t % polymer. The layer thickness was tuned to such a level that the absorbance in the UV-vis region did not exceed the value of 1 in order to obey the law of Lambert and Beer, which is necessary to justify further kinetic analysis. The layer thickness was estimated by dissolving a spin-coated polymer film in a known amount of solvent. The concentration of the azo groups could be found by of the solution with polymer comparingthe absorbance at )&,A solutions of known concentration. After estimation of the polymer density (.=1.5 g - ~ m -in~ )the film and measurement of the dimensions of the substrate, an estimation of the layer thickness could be obtained. The estimated thicknesses of the polymer films were on the order of 150 (3tZ.O) nm. Several substrates (glass, quartz, gold, or silicon) were used depending on the type of measurement. The UV spectra of all spin-coated films were blue-shifted compared to the UV spectra of the same
polymer in solution. The shift can be ascribed to an increased interaction between the Az.b. groups within the polymer matrix. Irradiation Apparatus. The film irradiation with UV was performed with a 1000 W xenon lamp combined with a monochromator and a shutter. The intensity was measured at 450 nm using a SpectrolineDRC-100Xradiometer. Filmswere irradiated The irradiated polymer at an intensity of 800 pW.cm-2at X,(t). films were transported in the dark to a separate Pye-Unicam SP8-200UV-vis spectrophotometer. The films were exposed to visible light to increase the rate of cis trans isomerization for “switching” experiments using a 15 W light bulb placed at a distance of 3 cm from the film, Itslight intensitywas 200pW-cm-2 at 450 nm and 30 pW.cm-2 at 360 nm. UV Measurements. A Pye-Unicam SP8-200 UV-vis spectrophotometer connected to a microcomputerwas used to record the UV spectra as a function of time. The substrates for the films were glass slides (9 X 30 X 1 mm) for polymers I, 11,111, and V and quartz (same dimensions) for polymers IV, VI, and VII. IR Measurements. Infrared measurements were performed with a Bruker IFS88 FT-IR spectrophotometer equipped with an MCT-A detector. A germanium Brewster angle IR polarizer was used for both the grazing incidence reflection (GIR) and transmission experiments. GIR spectra were recorded in an 80° specular setup with light polarized perpendicular to the surface. Films were spin-coated onto a gold layer; a clean gold layer was used as a reference. Transmission spectra were recorded from films spin-coated onto a silicium substrate using 10 cycles of 1000 scans according to the method of Arndt.13 The spectra were baseline corrected. Both GIR and transmission FT-IR have been used to determine the orientation of the Az.b. side groups in the LC polymers before and after UV and/or visible light exposure. Polarization Microscopy. The morphology of the spincoated films was studied with a Zeiss Axiophot polarization microscope with a camera mount and a Mettler FP-82 hot stage. The filmswere studied with crossedpolarizers to reveal LC phases. In this case the films were spin-coated from a 10 w t ’36 solution to obtain layers thick enough to be visible. Degradation Experiments. The degradation experiments were performed with a Karl Zeiss MA 56 apparatus at 365 nm with an intensity of 7.1 mW.cm-*.
-
3. Kinetic Analysis All measurements were performed at room temperature. The thermal isomerization rate constant k(Tisom)was determined by performing UV measurements as a function of time. The amount of trans Az.b. group is assumed to be linearly related with its absorbance at X,,(t) since during the whole process the absorbance was less than 1. Furthermore, the thermal cis trans back-isomerization is an irreversible process. By applying eq 1 k(Tbom)t can be computed. Plotting k( Tiaom)t versus time t gives a curve with slope k(TiBom).
-
Atransis the absorbance reached at infinite time, which is (normally) also the starting situation after storage in the dark before irradiation with UV light. A,& is the absorbance measured directly after irradiation, At is the absorbance at time t after irradiation. Values for the absorbance were taken at h,,(t). The half-life times t l p of the cis state were derived from absorption at t,,, = Acis+ (Atrans- A,,)/2
(2)
Since eq 3 assumes a first-order process and k(Tisom) was not found to be constant during the process of thermal back-isomerization, t 1 p was determined from eq 2.
Reversible Photoisomerization in Acrylic Polymers 6203
Macromolecules, Vol. 27, No. 21, 1994
0.60,
I
I
,
I;
0.40 I
250
300
-
400
350
500
450
550
wavelength (nm)A
Figure 3. Cis trans thermal back-isomerization of polymer 11. Curve A is trans before UV exposure; curve B is cis and trans in equilibrium directly after UV exposure. The arrows denote the direction of the thermal back-isomerization,and the dashed lines denote the isosbestic points. 4,
X
Y
n
q
N(cH,),
o
1
-
’i
I
3.
I
H
I
time xl~-’(min.)--
Figure 4. Plots of k( Tb,)t vs time t for the amorphous polymers.
4. Results and Discussion 4.1. Amorphous Polymers. Amorphous polymers I, 11,111,IV, and VI were irradiated by the UV xenon lamp at X,,(t) (see Table 1) during 25 min a t an intensity of 800 pW/cm2. The thermal back-isomerization process from cis to trans for polymer I1 (Y= N(CH3)2;X = H; n,m = 0; p/q = 0.510.5) is shown in Figure 3. The isosbestic points (denoted by dashed lines) a t 485 and 360 nm are characteristic for the existence of two distinct absorbing species in equilibrium with each other. The results for the kinetic analysis (tip, see Experimental Section) are gathered in Table 1 and Figure 4. As indicated in Figure 4,the k(Ti,,,)t versus t relation is not linear, as would be expected for first-order kinetics. This nonlinearity is thought to be caused by the difference in the environment of the azobenzene group^.^ The rate of the thermal isomerization is enhanced by steric factors such as limited free volume in the polymer matrix or (movements of) neighboring chain segments. The slope of k(Tisom)versus t can be regarded as a summation of individual k values, in which in the beginning fast azo groups enhance the overall rate and only slow groups remain after a period of time. This makes it impossible to give one absolute value for k(TiSom). The various (coipolymers can be compared better by using the half-life times of the cis state (tip), as presented in Table 1. It can be seen that a broad range of stability of the cis can be reached (tip ranges from 60 to 2700 min) by varying the Y substituent, spacer length, and copolymer composition. Copolymerization with acrylate, Le., “dilution” of the azobenzene group, has only a slight effect on the rate of
0.35-
0
1
I
I
,
1.5
3.0
4.5
6.0
1.5
time x1~-3(min.)-
Figure 5. Repeated UV/dark cycles with polymer 11. Curves A, B, and C show the thermal back-isomerization directly after UV exposure. the thermal isomerization (polymers I and 11, Table 1). The introduction of a spacer (polymer 111,Table 1)results in a decrease of t l p by a factor of 2 due to the increased floppiness of the azobenzene group. The variation of the substituent has the largest effect on the rate of the thermal back-isomerization. Changing from a N(CH& substituent4polymer I) to COzEt (polymer IV) causes an increase of t l p by a factor of 22, which is caused by an increase of the activation energy for the thermal isomerization due to a smaller overlap between the x-ir* and the n-x* states. This difference in overlap can be attributed to the electronegativity difference between the substituents of the Az.b. side group.’ The N(CH& group is a strong electron donor and induces an asymmetric electron distribution within the azobenzene group resulting in a push-pull type of azobenzene (large overlap), whereas in the case of the ester group a more symmetric electron distribution is found (small overlap). The photoisomerization and the thermal back-isomerization in the dark can be repeated many times, demonstrating its reversibility. Figure 5 illustrates the At of polymer I1 as a function of time in three UV/dark cycles. Each time the same starting absorption and a similar thermal back-isomerization process (curves A, B, and C) could be found. This process was repeated over 15 cycles without any signs of fatigue. The relation between the reached cis concentration in the photoisomerization and the duration of the exposure time for polymer I1 (Y = N(CH&; X = H; n,m = 0; p/q = 0.5/0.5;Figure 6 ) indicates that exposure times of about 1min are sufficient to reach a photostationary state with the light intensities used. This could be shortened by applying a higher light intensity. Compared to similar experiments for the same polymer I1 in solution,14 only half of the total trans azobenzene groups could be converted into cis in a thin solid polymer film. Apparently not a1lAz.b. side groups in the polymer matrix are available for photoisomerization, which is probably mainly due to steric hindrance. None of the polymers showed any signs of degradation during the repeated UV/dark cycles at the intensities used (800 pW-cm-Z), which was established with several spectroscopic techniques. To test the stability, degradation experiments were performed with 365 nm UV light with a very high intensity of 7.1 mW.cm-2. These showed no change in the UV absorption spectra within 30 s. After about 5 min of irradiation the absorption in the whole UV spectrum had decreased by a few percent, but still no
Macromolecules, Vol. 27, No. 21, 1994
Haitjema et al.
6204
k
04
f e
W-hP
hghtbulb
1
54 0.3c 9
02t
e
C